Auckland transport planners are advocating a $2.4 billion spend on an inner-city rail link which they believe will begin to ease traffic jams in the city. (The City Rail Link is underground rail tunnel that will run from Britomart, parallel to Queen Street, with stations at Aotea Square, K Road and in Newton. It is the first step in a system which will extend to the North Shore and to the airport. Effectively it will link Britomart to the Western line. It won’t affect transport on either the Southern motorway or from the North Shore)
So far the major arguments over this project involve the cost and the number of people who will use it. Can we believe that the costs won’t blow out in the way of most major infrastructure projects, and who will actually pay for it. And patronage is an issue. The government wants to delay the start until rail patronage reaches 20 million passengers (it is currently about half that) and until there is a 25% increase in people working in the CBD. This is not likely to happen before 2020.
Mayor Len Brown said a 2016 start to building the rail link was vital to Auckland’s economic development and would help to prevent crippling peak-hour road congestion. “Without the [Central Rail Link], by 2021 Auckland’s bus network will have reached capacity, and speeds on city roads will have dropped to a creeping 7kmh during peak time,” he said.
But there is a third factor which no-one seems to have considered – Google’s self-driving cars.
Google already has computer-controlled cars operating on American roads. These are standard cars adapted to computer control, but with a human driver ready to take over if necessary. They are legal and have racked up millions of miles safely. But these cars are more proof-of-concept than a game-changer as they are very expensive and don’t particularly improve road use efficiency.
The game-changer is Google’s latest version – a glorified golf cart with no steering wheel.
In theory, these cars will pack roads far more efficiently. They can travel nose-to-tail, and can put 3 cars abreast where today a road has 2 lanes. And they will improve your commute because you can read sleep or use your cell phone while you travel.
In practice, how likely is this to happen?
It took Google about 7 years to develop the first type of self-drive cars – from 2004 DARPA challenge to 2011 to get Nevada to license them for use on the road. That was basically developing the technology from scratch. The technology for the new car is already in place.
Acceptance will also be faster than many would anticipate.
Insurance companies already push drivers to greater supervision. In some places young drivers can only get (affordable) insurance if they install a system to monitor their driving, and the premium changes according to their scores. And if automated cars reduce accident claims, insurers will incentivise drivers to use them.
Councils and governments will see the benefits of being able to pack 4 or 5 times as many cars onto existing roads, because it means they won’t have to build more, so they will regulate to favour them. To start with they will ban all non-automated vehicles from major roads during rush hours, then this will be extended to daylight hours, and finally 24/7, by which stage owning a non-automated car will no longer be practical. And the CBD will no longer need to provide parking for cars (they can drive themselves to a temporary storage location when not needed) so more traffic can be accommodated or cycle lanes or footpaths extended.
And Joe Citizen will see the advantages. Get rid of your car and the expense of purchase, depreciation, insurance, repairs and maintenance and licensing. Commute to work without the stress of driving – the car can be an office or a lounge. Kids can be dispatched to school, grandma can be taken to the doctor.
The bottom line is that these cars could be ubiquitous by 2020. If that happens, rail patronage numbers will plummet, and the City Link will become an expensive underused white elephant.